I have a lot of interest in writing systems, which is why you will find a lot of posts on this blog around them. A few years ago, I wrote about the Bharati script, which was touted as an attempt to create a universal script for Indian languages. I had expressed measured skepticism about the idea, but had also said I would like to see the script, and luckily, was contacted by Chetan Shenoy, an undergrad from IIT Madras, who works under Prof Chakravarthy, the creator of the script. He let me know that the script had now been made public, and can be accessed here (this whole thing happened two years ago but I never ended up writing about it). I would first discuss the script a little bit, and then move on to discussing whether we really need a common script—if you are not interested in the specifics of the script, skip to the next section.

The Bharati script is a motley combination of symbols largely from Devanagari, Tamil and the Roman script. Prof Chakravarthy had expressed disappointment, and I had agreed [to the facts, not the inference], that despite the scientifically structured ordering of the symbols in Indian scripts, [which shows that the scholars of the past did have cognition of the relevant linguistic facts], the shapes of the letters were unnecessarily arbitrary. This makes sense, because the writing system is more likely to have developed organically, and would have preceded scholarship about the same.

In order to rectify this, the Bharati script has a standard mark for aspiration (a dot) and for voicing (a vertical line) that accompanies the symbol of the base form (unaspirated, unvoiced). The root letter for K series is taken from the Devanagari script (minus the vertical bar). Likewise the root letter for the CH row. From here, we move on to Tamil, with the letter for the T row sourced directly from the script used to write Tamil. The next row again moves back to Devanagari (albeit with the vertical line in this case). The root for the fifth row (P) is from Tamil. The corresponding nasals are a bit funny. N and M have been sourced from the Roman script, while the retroflex N is basically the corresponding Devanagari character. The velar and palatal nasals, I can’t associate with any Indian script of the top of my head. Y, R and V and S are their English counterparts, while the L is sourced from Kannada. Conspicuously absent is the symbol corresponding to the Tamil ழ or Malayalam ഴ, so as it stands, the name of the Tamil language (தமிழ்) can’t be written using this script. I found that a bit surprising, especially as the professor is from IIT Madras.

I didn’t quite like Bharati and could go on a bit more about it but I think that misses the bigger point. As I see it, there are three questions here to consider—How much does it matter if the letters of a script are logical? Does India need a new common script? Finally, if we do need one, what should that common script be?

Let’s ignore logographic writing systems such as the Chinese Hanzi, or the Japanese Kanji that derives from it. This is not merely a convenient omission, it’s a reasonable one, for if your focus is on developing a new script for easy universal adoption, the Hanzi is the last place you’d want to look.

Ignoring Chinese-like writing systems, if you look at prevalent writing systems around the world, you’d notice that they are fairly easy to learn. Though they are of a few different types, most have fewer than 100 characters, and the strong relation between sound and writing in most of them make them fairly easy to pick up if you learn the script along with learning the language. Note here that I am talking specifically about the script, not about the spelling system of the language. These differ, and the same script can be used by different languages in different ways. The best example of this is of course the Latin Script, which is used by more than 100 languages. The English spelling system, for instance, is notoriously irregular and difficult to navigate the intricacies and idiosyncrasies of, as I have discussed on several occasions on this very blog. Spanish, on the other hand, uses the same script, but has a very predictable and regular spelling system, and you can learn to pronounce Spanish words correctly with very little study.

Closer home, most Indian scripts are also fairly straight-forward, and while we have a lot of them, they are mostly all arranged similarly, and if you really wanted, you could learn all the symbols of a particular script in a few days. If you are actually learning the language along with it, you will retain the knowledge much better and would be able to read faster. Some scripts are extremely easy, such as Tamil, but that is because it has too few characters, and ends up merging all voicing and aspiration distinctions into one, while others are a bit difficult1.

All these writing systems are relatively easy to pick up, but the shapes of characters in most of them are _not_ motivated by phonetic analyses, and the role phonetics plays in most of them is pretty small, if any. Some do mark phonetic features uniformly2, but overall, it’s still fairly arbitrary.

My point is that while regularity in a script’s characters is a good-to-have (if you really care about it), it’s not a need-to-have when the number of characters is relatively low (less than 100), as is the case with most prevalent writing systems in the world. Given the degree to which these systems are entrenched into everyday life, the motivation for regularizing what’s already working well is akin to fixing what ain’t broke.

This brings us to the next question. Does India need a common script? Everything else held equal, having a common writing system is of course trivially better than having different writing systems, just like having a language everyone knows is trivially better than having no common language.English-literate people can trivially hop on to learn, say, French, or German, or Malay, for that matter, without having to learn a new script. As I have argued above, learning a script is not that huge of a barrier, but having no barrier is trivially better than having some barrier. I have argued in a previous post about how the lack of a common language in India comes with its own set of issues which can’t just be romanticized away by diversity arguments.

But that is after ‘holding everything else equal’. People have strong emotional ties with their writing system, scripts are a huge part of the culture of the people who speak it, and then of course there are huge financial costs of changing your entire system to a new one. So, if you are doing it, you better have very very good reason. And for all the millions who are already using the old script, there is scarcely any reason to want to move to a new one3.

If you really really want to switch to a common script for all Indian languages, switch to the Latin one. Yes, I know it’s rather drastic, but hear me out. If you are willing to go through, and subject everyone in the country to all this pain just to shift to a new script, you might as well get a LOT of bonus features along the way.

  • ‘Ambiguity of Latin script’ is an irrational bugaboo, as I have explained above. Romanization systems for Indian languages are already pretty regular to begin with, and if you are planning to shift, can be standardized completely. It might have to be slightly different for different Indian languages, but that problem will exist with any common script.
  • The whole argument about being sentimental about your own script and culture was presumably thrown out of the window anyway when you chose to adopt a universal script. And if you are going to run it over, might as well use the opportunity to adapt the closest thing we have to a global adoption standard.
  • The phonetically motivated ordering of letters in Indian scripts is entirely irrelevant to how the script is used, so nothing is lost in this conversion.
  • Indian texts will become much more accessible on the internet, much more easily searchable, your languages becomes more accessible to outsiders and to insiders, you get a largely unified way of spelling things within the country, you end up learning the Latin script for free, so it’s easier for people to learn English if they want to. I mean, the possibilities are endless.
  • A bonus with choosing Latin of course is that it helps you bypass all Hindi-Tamil-Whatever squabbles. Even today, it’s English that is India’s unifying language.

To clarify, I am not saying we must change to the Latin script, but that if we are planning to move the country to a common script, we might as well make it Latin. It’s a defensible proposition, and worth all the trouble that any such process will entail. And if you are lamenting losing all the beautiful Indian scripts in the process, rest assured that Bharati or any similar attempt is not going to be any different in that regard.

  1. Such as Urdu with the do-chashmi-he for aspiration, Kannada / Telugu with the dot for aspiration (but only in some cases), Hangeul (Korean) with the extra hyphen-esque line for aspiration (in consonants) and for adding a y sound to vowels. Now that I think of it, aspiration seems to be a popular thing to mark using such techniques. [return]
  2. Bengali tends to go overboard with its conjunct characters (যুক্তাক্ষর, juktakkhôr), but it’s tractable nonetheless. [return]
  3. One of the most famous examples of a culture adopting a new writing system is that of the adoption of Hangeul, the script Korean is written in. But, Hangeul was specifically created because Korean, until then, was written using Chinese characters, which were not accessible to the general population. Hangeul was supposed to be a replacement so that literacy was accessible to everyone, not just the intellectual class. [return]